A major goal of this program is to improve the sensitivity and specificity of diagnosis in the common pediatric tumors through application of molecular biological techniques. An additional goal is to use pediatric tumor specimens to define molecular process which contribute to the malignant phenotype of childhood solid tumors. Common childhood tumors that will be studied include acute lymphoblastic leukemia, Wilms tumor, neuroblastoma, medulloblastoma, rhabdomyosarcoma and osteogenic sarcoma. Project 1 will utilize specific oligonucleotide probes, which recognize the CDR3 region of the IgH locus of the leukemic clone, to quantify the number of leukemic cells in the bone marrow and peripheral blood of patients with subclinical disease. It will be learned if the presence of large numbers of leukemic cells are predictive of relapse after induction- or during maintenance-chemotherapy. Project 2 characterizes the Wilms tumor zinc finger protein in tumor samples by the ability to bind to a specific DNA sequence and regulate transcription. Project 3 will identify and characterize NGF induced immediate-early genes in primitive neuroectodermal tumors which are incompetent to trans activity other genes whose expression is required for terminal differentiation. Project 4 identifies and clones the t(2;13) translocation breakpoint characteristic of alveolar rhabdomyosarcoma because this translocation likely contributes to tumorigenicity. Project 5 identifies osteosarcomas which produce PDGF and learns if endogenous PDGF production causes osteosarcomas to grow in an autocrine fashion. It also identifies genes and transduction mechanisms which are impaired in those osteosarcomas which don't respond to PDGF.